Genetic polymorphism of rs9564966 G > A on 13q22.1 predicts poor survival for Chinese patients with gastric cancer

Abstract Two genomewide association studies on pancreatic cancer have identified a novel single‐nucleotide polymorphism of rs9564966 G > A on 13q22.1 region. However, the associations between the rs9564966 G > A polymorphism and the survival of Chinese patients with gastric cancer (GC) were unknown. In our present investigation, we adopted the Kaplan‐Meier plots, Cox regression analyses, and the log‐rank tests to explore the associations between rs9564966 G > A polymorphism and the prognosis of 911 Chinese patients with GC. Our results revealed that, compared with GG genotype, patients with GA + AA genotypes had poorer outcomes (HR = 1.348, 95% CI = 1.084‐1.675, P = 0.007), especially in the subgroups of age ≤60 years, male, nondrinker, tumor size >5 cm, tumor site in Noncardia, intestinal‐type tumor, T3/T4 level depth of invasion, N1/N2/N3 level lymph node metastasis, no distant metastasis, III/IV level TNM stages, and no chemotherapy. Our findings suggested that the rs9564966 G > A polymorphism may be a potential biomarker to predict the survival of Chinese patients with GC.

Due to Rs9564966 was located between KLF5 and KLF12 on chr 13q22.1, so it may also play essential roles in the regulations of various cellular processes by influencing the functions of KLF5 and KLF12. The abrogation of genes located on chromosome 13 was related to various cancers, including pancreatic, prostate, and breast cancers. [19][20][21] Some studies reported that KLF5 acted as tumor suppressor to inhibit cell proliferation and function; KLF5 gene underwent frequent alteration and expression aberration in various cancer types. [22][23][24][25][26] Other studies demonstrated that KLF12 knockdown significantly inhibited cell proliferation and overexpression of KLF12 promoted cell invasion. 11 The levels of KLF12 mRNAs and proteins were obviously unregulated for cancer patients, revealing that KLF12 may confer growth advantage on cancer cells and play crucial roles in tumor progression. 11 As a novel locus on 13q22.1 located between KLF5 and KLF12, whether SNP rs9564966 was associated with the survival of Chinese patients with GC was unknown. Therefore, we performed this study to explore the associations between rs9564966 polymorphism and the clinical prognosis of Chinese patients with GC.

| Study patients
All 919 GC patients who underwent surgery at the Yixing People's Hospital (Yixing, Jiangsu Province, China) between January 1999 and December 2006 were enrolled in our study. All the patients were histopathologically diagnosed and had no chemotherapy or radiotherapy before surgery. The patients' clinical features were shown in Table 1. The survival time of GC patients was calculated from the date of surgery until death or to the last follow-up in March 2009, and the median follow-up time was 68.5 months. Tumors were classified into intestinal-type or diffuse-type by Lauren's criteria. 27 The TNM stages were evaluated according to the TNM classification of the American Joint Committee on Cancer (AJCC cancer staging manual, seventh edition). All patients recruited in our study signed an informed consent on the use of clinical specimen for medical research. The institutional review board of Nanjing Medical University approved our present study. And all methods included in our investigation were performed in accordance with the relevant guidelines and regulations.

| Genotyping
Genomic DNA was extracted from paraffin sections of patients' postoperative tissues by proteinase K digestion, isopropanol extraction, and ethanol precipitation. 28 Rs9564966 G > A polymorphism was detected by SNaPshot technology based on an ABI fluorescence assay allelic discrimination method (Applied Biosystems, Forster City, CA). We designed the primers to anneal immediately adjacent to the nucleotide at the mutation site: rs9564966: forward, 5′-GAGAAAAGCCAGTTACATTACAGAACTTCC -3′; reverse, 5′-GAAAGATCACTAGGGCCCCTTCC -3′.
The primers for extension were as follow: rs9564966, 5′-TTTTTGCCAGTTACATTACAGAACTTCCTTGATG-3′. ABI3130 genetic analyzer was used to analyze the SNPs and the Genemapper 4.0 software (Applied Biosystems) was used to determine the genotypes. Genotyping assays were performed by two investigators independently in a blind fashion. Ten percent samples were randomly selected to validate genotypes and the results were 100% concordant.

| Statistical analysis
The Student's t test for continuous data and the Pearson's chi-square test for categorical variables were employed in estimating the associations between rs9564966 polymorphism and clinicopathologic parameters. When the median survival time (MST) could not be calculated, the mean survival time was chosen. Kaplan-Meier plots and log-rank tests were analysis by SPSS version 20.0 (SPSS Inc, Chicago, IL, USA). Univariate or multivariate Cox regression analysis was used to calculate the crude or adjusted hazard ratios (HRs) and the 95% confidence intervals (CIs). Moreover, we performed Cox stepwise regression analysis to assess the independent impacts of SNP or clinicopathologic features on the overall survival, with a significance level of P < 0.05 for entering and P > 0.10 for removal of the respective explanatory variables. All tests were two-sided and P < 0.05 was considered statistically significant.

| Clinical characteristics of the study patients
A total of 919 patients were recruited in our investigation, and eight patients were removed due to lack of genotyping data. Finally, 911 patients with GC were enrolled and the clinical features of those patients were summarized in Table 1. All of the study subjects underwent surgery, and the median age was 62.0 years (range from 28 to 83 years old). Two hundred and eleven women and 700 men were included and 418 of them died during the follow-up period of 119 months. Significant associations were founded between the survival time and the varieties in tumor size, histological types, depth of invasion, lymph node metastasis, and TNM stage (log-rank < 0.001). In particular, 95% CI = 1.435-2.182) compared with those without metastasis (MST, 81 months). In addition, as the TNM stage increased, the risk of death also exhibited a significant increase (log-rank P < 0.001).

| The polymorphism of rs9564966 located on 13q22.1 increased death risk
Cox regression analyses were applied to analysis the relationships between rs9564966 polymorphisms and the survival of patients with GC in different genetic models, including codominant, dominant, and recessive models ( Table 2). Our results firstly demonstrated that rs9564966 G > A polymorphism obviously increased the death risk of patients with GC. Compared with GG genotype, rs9564966 GA genotype significantly increased the death risks of patients with GC (HR = 1.444, 95% CI = 1.152-1.811, P = 0.003, Figure 1A). Similarly, the MST of GC patients with GA + AA genotypes was shortened from 98 to 56 months (HR = 1.348, 95% CI = 1.084-1.675, P = 0.007, Figure 1B) when compared to those with GG genotype.  Figure 2), and no chemotherapy (HR = 1.308, 95% CI = 1.005-1.703, P = 0.044).

| The independent impacts of SNP or clinicopathologic features on the overall survival after adjusting for other covariates
In the end, stepwise Cox regression analysis was used to analysis the associations between the rs9564966 G > A polymorphism, patients' clinical features, demographic characteristics, and GC patients' survival. As shown in Table 4, the rs9564966 G > A polymorphism and varieties in tumor size, lymph node metastasis, and depth of invasion were significantly related with the patient survival rates: P = 0.003 for rs9564966 G > A

| DISCUSSION
In our present study, we firstly investigated the clinical effects of rs9564966 G > A polymorphism on the prognosis of Chinese patients with GC. The dominant genotypes GA + AA were obviously related with poor survival of patients with GC, especially in the subgroups of age ≤60 years, male, nondrinker, tumor size >5 cm, tumor site in noncardia, intestinal-type tumor, T3/T4 level depth of invasion, N1/N2/ N3 level lymph node metastasis, no distant metastasis, III/IV level TNM stages, and no chemotherapy. Two GWASs on pancreatic cancer have discovered a novel SNP rs9564966 G > A in gene deserts of 13q22.1 region, which may act as a blasting fuse disrupting the function of nearby normal gene through different mechanisms. [12][13][14] SNP rs9564966 was located on the first sub-band region of 13q22 between transcription factors KLF5 and KLF12. 10,11 KLF5 was associated with numerous transcription factors and played vital roles in reducing development, proliferation, and transformation of different cells. 16,17 And KLF12 suppressed the expression of the AP-2a transcription factor through binding to the promoter region and overexpression of KLF12 promoted cell invasion. 11,18 So, the polymorphism of rs9564966 G > A on 13q22.1 region may also play a vital role in the initiation and progression of GC by influencing the functions of KLF5 and KLF12.
Several studies showed that the expression of KLF5 was downregulated or absent in patients with GC, 29 and similar reduction was also occurred in other cancers, including prostate, breast, esophageal, and intestinal cancer. 22,24,25 Nakamura et al 11 showed that KLF12 level was remarkably upregulated for patients with GC and the knockdown of KLF12 induced growth arrest and significantly inhibited GC cell proliferation and invasion. For the SNP of rs9564966 G > A, recent studies demonstrated that the polymorphism of rs9564966 G > A significantly increased the risk of pancreatic cancer, 8,9,19 which was similar to our results in patients with GC. Our finding indicated that patients with GA + AA genotypes had poorer prognosis than those with GG genotype (HR = 1.348, 95% CI = 1.084-1.675, P = 0.007, Figure 1). The explanation for this fact may be that the SNP of rs9564966 G > A affected the expressions and functions of the nearest upstream and downstream genes, including KLF5 and KLF12. 3 The polymorphism of rs9564966 G > A may reduce KLF5 expression and improve KLF12 level through different mechanisms, exerting vital roles in tumor progression. Reduced expression of KLF5 would decrease the growth inhibitory function and the improved expression of KLF12 could promote cell proliferation and invasion. 11,29 Stratified analysis of the SNP with various subgroups based on clinical features was vital to identify the potential prognostic markers for patients with GC. Our findings observed that, compared with the GG genotype, the rs9564966 GA + AA genotypes could significantly decrease the survival time of GC patients with age ≤60 years, male, nondrinker, tumor size >5 cm, intestinal type of cancer, T3/ T4 level depth of invasion, N1/N2/N3 level lymph node metastasis, no distant metastasis, and III/IV level TNM stage.

T A B L E 4
Stepwise Cox regression analysis on the survival of patients with gastric cancer Furthermore, GA + AA patients without chemotherapy had poorer prognosis than those with GG genotype, and no similar result was found among patients underwent chemotherapy. The expression changes of KLF5 and KLF12 genes affected by rs9564966 polymorphism may be also the possible explanations of this phenomenon. [12][13][14] Nevertheless, several limitations were existed in our present study. First, we only examined SNP of rs9564966 located on 13q22.1, but more polymorphisms at other loci and the synergistic effects of those genes interacted with each other should be further tested. Second, the maximum follow-up time of the patients recruited in our study was 119.0 months, so more studies with longer follow-up time were warranted to be carried out to validate our results. Third, our present study was carried out in Chinese, whether similar results also existed among different populations should be tested in other researches. Last, H. pylori infection played a crucial role in the process of gastric carcinogenesis. But, the status of H. pylori infection was not included in our study for lack of clinical information.
In summary, we observed that rs9564966 G > A polymorphism was associated with the poor survival of patients with GC, suggesting that the polymorphism may be a useful potential biomarker for the prognosis of Chinese patients with GC.